Dell Acceleration Appliance for Databases vs. IBM ... › Dell › DAAD_vs_IBM_0815.pdfDell Acceleration Appliance for Databases vs. IBM FlashSystem 900: Capacity, cost, and performance

DELL ACCELERATION APPLIANCE FOR DATABASES VS. IBM FLASHSYSTEM 900: CAPACITY, COST, AND PERFORMANCE

MAY 2015

A PRINCIPLED TECHNOLOGIES REPORT Commissioned by Dell Inc.

Even with powerful database servers, it’s important to give your business every

possible advantage to keep up with the growing numbers of users and applications

accessing your company’s resources. Upgrading your storage infrastructure for your

database applications with the highly available Dell Acceleration Appliance for

Databases (DAAD) 2.0 powered by Intel Xeon processors E5-2667 v3 – a high-

performance flash storage solution – can help you achieve this goal.

In the Principled Technologies datacenter, we compared the Oracle Database

12c performance of two environments: one with a highly available Dell Acceleration

Appliance for Databases (DAAD 2.0 solution) and one using an IBM FlashSystem 900

storage appliance (IBM solution). We found that the highly available DAAD solution

comes at a 37.0 percent lower cost per terabyte of storage space compared to the IBM

FlashSystem 900 appliance, so you get more storage capacity for your dollar. The DAAD

solution also provided 2.9 percent better transactional database performance than the

IBM solution, while offering critical redundancy to reduce the possibility of devastating

downtime in case of a failure. According to IBM, FlashSystem 900 does not come in a

highly available configuration, which could lead to lost revenue and more importantly

lost data, if a failure happens. If you’re looking to maximize database performance and

optimize storage capacity per dollar, the Dell Acceleration Appliance for Databases can

meet your needs.

http://www.principledtechnologies.com/

A Principled Technologies report 2

Dell Acceleration Appliance for Databases vs. IBM FlashSystem 900: Capacity, cost, and performance

HOW IT WORKS The Dell Acceleration Appliance for Databases 2.0 is a pre-integrated

server/software combination consisting of one or two Dell PowerEdge R730 servers with

Intel Xeon processors E5-2667 v3, DAAD ION Accelerator software, and Fusion SX300

ioMemory drives, which are high-capacity NAND Flash PCIe SSDs. Available in 12.8TB

and 25.6TB capacities, DAAD 2.0 supports Fibre Channel, iSCSI, and Infiniband/SRP

front-end fabrics. Designed to work with many database platforms and front-end

servers, DAAD makes appliance resources available to the application server and uses

I/O acceleration to optimize database requests, leveraging DRAM to improve

performance. Boosting database performance with DAAD can help you work within your

existing infrastructure, so you don’t have to purchase more disk or compute resources,

incur additional database licensing costs, or take up vast amounts of extra datacenter

space. To learn more about the Dell Acceleration Appliance for Databases, visit

en.community.dell.com/techcenter/enterprise-

solutions/m/oracle_db_gallery/20441362/download.

In this study, we configured two 25.6 TB DAAD appliance nodes as an HA cluster

pair. For information about the DAAD and our other test components, see Appendix A.

For detailed system configuration, see Appendix B. For step-by-step testing details, see

Appendix C.

WHAT WE FOUND As Figure 1 shows, we found that the highly available DAAD solution cost 37.0

percent less per terabyte of storage. This means that with a fixed budget, you can have

more usable storage capacity to handle growing business needs. (See Figure 2 for

details.)

Figure 1: The highly available Dell Acceleration Appliance for Databases solution has a 37.0 percent lower cost per terabyte than the IBM FlashSystem 900 appliance.

http://en.community.dell.com/techcenter/enterprise-solutions/m/oracle_db_gallery/20441362/downloadhttp://en.community.dell.com/techcenter/enterprise-solutions/m/oracle_db_gallery/20441362/download



Usable storage capacity

(TB) Cost

(USD)1 Cost/terabyte

DAAD solution 25.62 $441,000.76 $17,226.59

IBM FlashSystem 900 7.2 $196,800.00 $27,333.33

Figure 2: Storage capacity and cost comparison for the two solutions.

In addition to more storage capacity and a lower cost per terabyte, we found

that the highly available Dell Acceleration Appliance for Databases with Intel Xeon

processors E5-2667 v3 also outperformed the IBM solution. The DAAD solution provided

2.9 percent better OLTP Oracle Database 12c performance than the IBM FlashSystem

900 solution (see Figure 3). In our tests, we ran a TPC-C-derived workload and used

HammerDB.

Figure 3: Using the Dell Acceleration Appliance for Databases solution provided 2.9 percent better database performance. The results we report are from the median of three runs.

When you’re deciding which flash storage solution to buy to upgrade your

storage infrastructure for your database applications, your goal is to accommodate

existing users while leaving headroom for expanding needs as your business grows.

With a DAAD, you get a performance boost for your applications, as well as the

protection that a highly available configuration offers in the event of a failure.

1 DAAD solution price was provided by Dell as of 6/2/2015 and IBM FlashSystem900 price is from an IBM reseller as of 6/2/2015. Prices are list prices and do not include discounts, tax, or shipping costs. 2 A two-node, highly available DAAD configuration provided the 25.6 TB usable capacity.



CONCLUSION There’s no better time to upgrade your database environment than when your

business is expanding and you need more performance to keep up with growing

demand. Flash storage is an attractive solution because it provides significant

performance boosts over spinning disks and traditional storage arrays, but selecting the

fastest solution with the best storage capacity isn’t always easy. We found that a highly

available Dell Acceleration Appliance for Databases solution provided 2.9 percent better

database performance than the IBM FlashSystem 900, plus 37.0 percent lower cost per

terabyte of storage capacity, which resulted in incredible performance at an efficient

price per TB.



APPENDIX A – ABOUT THE COMPONENTS About the Dell Acceleration Appliance for Databases 2.0 hardware

The two 2U Dell PowerEdge R730 rack servers used in the highly available DAAD configuration are each powered

by two Intel Xeon processors E5-2667 v3, 384 GB of DDR4 RAM, and QLogic QLE2662 16Gb Fibre Channel HBAs,

providing functional flexibility in the datacenter. The Dell PowerEdge R730 rack servers are each powered by four 6400

GB SanDisk SX300 ioMemory PCIe® SSDs to reduce storage bottlenecks.

With redundant power supply units, hot-swappable hardware, and Dual SDTM card for Failsafe Hypervisors, the

Dell PowerEdge R730 supports hardware high availability. The PowerEdge R730 comes standard with iDRAC8 with

Lifecycle Controller and Dell OpenManage, which all work to streamline management. For more details on the Dell

PowerEdge R730, visit www.dell.com/us/business/p/poweredge-r730/pd?oc=pe_r730_1356&model_id=poweredge-

r730&l=en&s=bsd. For more details on the Intel Xeon processor E5-2600 v3 series, visit

www.intel.com/content/dam/www/public/us/en/documents/product-briefs/xeon-e5-brief.pdf.

About the benchmark we used - HammerDB HammerDB is an open-source benchmark tool that tests the database performance of many databases,

including Oracle Database, Microsoft® SQL Server®, PostgreSQL®, MySQL™, and more. The benchmark includes built-in

workloads derived from industry-standard benchmarks, such as a transactional (TPC-C-derived) workload and a data

warehouse (TPC-H-like) workload. For this study, we used the transactional workload. The TPC-C HammerDB workload is

derived from TPC-C and as such is not comparable to published TPC-C results. For more information, visit

www.hammerora.sourceforge.net.

http://www.dell.com/us/business/p/poweredge-r730/pd?oc=pe_r730_1356&model_id=poweredge-r730&l=en&s=bsdhttp://www.dell.com/us/business/p/poweredge-r730/pd?oc=pe_r730_1356&model_id=poweredge-r730&l=en&s=bsdhttp://www.intel.com/content/dam/www/public/us/en/documents/product-briefs/xeon-e5-brief.pdfhttp://www.hammerora.sourceforge.net/



APPENDIX B – SYSTEM CONFIGURATION INFORMATION Figure 4 provides detailed configuration information for the test systems.

System Dell PowerEdge R720 (Oracle

Database server)

Dell Acceleration Appliance for

Databases 2.0

Power supplies

Total number 1 2 (HA pair)

Vendor and model number Dell 05NF18X02 Dell 0G6W6KX02

Wattage of each (W) 750 750

General

Number of processor packages 2 2

Number of cores per processor 8 8

Number of hardware threads per core 2 2

System power management policy Performance Performance

CPU

Vendor Intel Intel

Name Xeon Xeon

Model number E5-2650 v2 E5-2667 v3

Socket type LGA 2011 FCLGA2011-3

Core frequency (GHz) 2.6 3.2

Bus frequency 7.2 GT/s 9.6 GT/s

L1 cache 32 + 32 KB (per core) 32 + 32 KB (per core)

L2 cache 256 KB (per core) 256 KB (per core)

L3 cache 20 MB 20 MB

Platform

Vendor and model number Dell PowerEdge R720 Dell PowerEdge R730

Motherboard model number 0M1GCR 0599V5

BIOS name and version 2.4.3 1.1.4

BIOS settings Defaults Defaults

Memory module(s)

Total RAM in system (GB) 128 384

Vendor and model number Hynix HMT31GR7BFR4A-H9 Hynix HMA42GR7MFR4N-TFT1

Type PC3-10600 PC4-17000

Speed (MHz) 1,333 2,133

Speed running in the system (MHz) 1,333 2,133

Size (GB) 8 16

Number of RAM module(s) 16 24

Chip organization Double-sided Double-sided

Rank 2Rx4 2Rx4

Operating system

Name Red Hat® Enterprise Linux® DAAD ION Accelerator

Build number 6.6 2.5.1-413

File system ext4 btrfs



System Dell PowerEdge R720 (Oracle

Database server)

Dell Acceleration Appliance for

Databases 2.0

Kernel 2.6.32-504.el6.x86_64 3.0.101-0.15.1.6651.0.PTF-

default (x86_64)

Language English English

RAID controller

Vendor and model number Dell PERC H710P Mini Dell PERC H730P Mini

Firmware version 21.3.0-0009 25.2.1.0037

Cache size (GB) 1 2

RAID configuration 1 x RAID50 1 x RAID1

Hard disk types

Hard disks (OS)

Vendor and model number Dell ST9146852SS Dell ST300MM0006

Number of disks 16 2

Size (GB) 146 300

RPM 15K 10K

Type SAS SAS

PCIe SSDs

Vendor and model number N/A Fusion ioMemory SX300

Number of disks N/A 4

Size (GB) N/A 6,400

Type N/A PCIe

Ethernet adapters

Vendor and model number Intel I350 Quad-port 1GbE Broadcom® NetXtreme®

BCM5720 Quad-port 1 GbE

Firmware 14.5.9 7.10.18

Type rNDC rNDC

Fibre Channel adapters

Vendor and model number QLogic® QLE2672 16Gb Fibre Channel

Adapter

QLogic QLE2672 16Gb Fibre

Channel adapter

Firmware 03.11.09 03.11.09

Type PCIe PCIe

USB ports

Number 4 4

Type USB 2.0 USB 2.0

Figure 4: Detailed information for the test storage.



Figure 5 shows the configuration information for the IBM FlashSystem 900 appliance.

Storage array IBM FlashSystem 900

Firmware revision 1.2.0.11

Disks – PCIe SSDs

Number of disks 6

Disk vendor and model number IBM MicroLatency Module

Disk size 1.2 TB

Figure 5: Detailed information for the test storage.



APPENDIX C – SETTING UP THE STORAGE Dell Acceleration Appliance for Databases solution

For the Dell solution, we replaced the IBM FlashSystem 900 with the highly available Dell Acceleration Appliance

for Databases, storing our data, logs, and backups on LUNs created on the DAAD. For our storage area network (SAN),

we attached the R720’s 16Gb Fibre Channel HBA as well as the DAAD’s 16Gb Fibre Channel HBAs to a Brocade SW6505

Fibre Channel switch.

Figure 6: The highly available DAAD configuration we tested.

IBM FlashSystem 900 solution For the IBM solution, we configured a Dell PowerEdge R720 server with Red Hat Enterprise Linux 6.6. We then

configured a standalone Oracle Database 12c Enterprise Edition installation with a 5,000-warehouse database. We used

Oracle Automatic Storage Management (ASM), storing our data, logs, and backups on LUNs created on the IBM

FlashSystem 900. For our storage area network (SAN), we attached the R720’s 16Gb Fibre Channel HBA as well as the

IBM FlashSystem 900’s 16Gb Fibre Channel HBAs to a Brocade SW6505 Fibre Channel switch.

Figure 7: The IBM FlashSystem 900 configuration we tested.



Setting up storage on the DAAD In a terminal, ssh into one of the DAAD nodes with the admin credentials.

Enter the following commands to create mirrored volumes across the DAAD nodes, an initiator group, and then

eight LUNs to be presented to the database server.

profile:create direct

volume:create -n ion01 -n ion02 fcion_v_a1 50% jbod_pool-1

volume:create -n ion02 -n ion01 fcion_v_a2 100% jbod_pool-1

volume:create -n ion01 -n ion02 fcion_v_b1 50% jbod_pool-2

volume:create -n ion02 -n ion01 fcion_v_b2 100% jbod_pool-2

volume:create -n ion01 -n ion02 fcion_v_c1 50% jbod_pool-3

volume:create -n ion02 -n ion01 fcion_v_c2 100% jbod_pool-3

volume:create -n ion01 -n ion02 fcion_v_d1 50% jbod_pool-4

volume:create -n ion02 -n ion01 fcion_v_d2 100% jbod_pool-4

inigroup:create ig_all 20:01:00:0e:1e:09:d6:1c 20:01:00:0e:1e:09:d6:1d

lun:create fcion_v_a1 ig_all -b 512 –a

lun:create fcion_v_a2 ig_all -b 512 -a

lun:create fcion_v_b1 ig_all -b 512 -a

lun:create fcion_v_b2 ig_all -b 512 -a

lun:create fcion_v_c1 ig_all -b 512 -a

lun:create fcion_v_c2 ig_all -b 512 -a

lun:create fcion_v_d1 ig_all -b 512 –a

lun:create fcion_v_d2 ig_all -b 512 -a

Setting up storage on the IBM FlashSystem 900 We performed the initial setup steps provided with the system documentation to connect it to our network.

Navigate to the IBM FlashSystem 900’s IP address in a browser, and log in with administrator credentials.

In the left dock, hover over the Volumes icon, and click Volumes.

Click Create Volumes.

Provide a name, size, and number of volumes.

We created two 1000 GB volumes named backup, four 300 GB volumes named data, and four 100 GB volumes

named logs

In the left dock, hover over the Hosts icon, and click Hosts.

Click Add Host.

Enter a name for the host. We used OracleR720

Click the drop-down menu next to Host port (WWPN), and select one of the ports.

Click the + button, and do the same with the other port.

Click Add.

In the left dock, hover over the Hosts icon, and select Volumes by Host.

Select all of the volumes in the Unmapped Volumes list, right-click them, and select Map to Host.

Make sure the desired host is selected, and click Map.



Configuring Red Hat Enterprise Linux and Oracle Database 12c We installed Red Hat Enterprise Linux on the Dell PowerEdge R720 server and configured settings as we specify

below. Screen outputs are in grey boxes.

Installing Red Hat Enterprise Linux 1. Insert the Red Hat Enterprise Linux 6.6 DVD into the server and boot to it.

Select Install or upgrade an existing system.

If you are unsure of the fidelity of your installation disk, select OK to test the installation media; otherwise, select

Skip.

In the opening splash screen, select Next.

Choose the language you wish to use, and click Next.

Select the keyboard layout, and click Next.

Select Basic Storage Devices, and click Next.

Click Yes, discard any data at the Storage Device Warning.

Insert your hostname, and click Next.

Select the nearest city in your time zone, and click Next.

Enter your root password, and click Next.

Select Create Custom Layout, and click Next.

Select the install drive and click Create. (Create the following volumes and size: Root = 300GB, Home = 500GB, Boot

= 200MB, SWAP = 20GB)

Click Next.

Click Write changes to disk.

Select the appropriate Data Store Devices and select where the Bootloader will go, and click Next.

Select Software Basic Server, and click Next. Linux installation begins.

When the installation completes, select Reboot to restart the server.

Performing initial configuration tasks Complete the following steps to provide the base functionality that Oracle Database requires. We performed all

of these tasks as root.

1. Disable SELINUX. vi /etc/selinux/config

SELINUX=disabled

2. Set CPU Governor type. vi /etc/sysconfig/cpuspeed

GOVERNOR=performance

3. Disable the firewall for IPv4 and IPv6. chkconfig iptables off

chkconfig ip6tables off

4. To update the operating system packages, type the following: yum update -y

5. To install additional packages, type the following commands:



yum install -y acpid cpuspeed wget vim nfs-utils openssh-clients man

lsscsi unzip smartmontools numactl ipmitool OpenIPMI

6. Reboot the server. reboot

7. Install additional packages with the following commands: yum install -y \

binutils \

compat-libcap1 \

compat-libstdc++-33 \

compat-libstdc++-33.i686 \

device-mapper-multipath \

gcc \

gcc-c++ \

glibc \

glibc.i686 \

glibc-devel \

glibc-devel.i686 \

ksh \

libgcc \

libgcc.i686 \

libstdc++ \

libstdc++.i686 \

libstdc++-devel \

libstdc++-devel.i686 \

libaio \

libaio.i686 \

libaio-devel \

libaio-devel.i686 \

libXext \

libXext.i686 \

libXtst \

libXtst.i686 \

libX11 \

libX11.i686 \

libXau \

libXau.i686 \

libxcb \

libxcb.i686 \

libXi \

libXi.i686 \

make \



sysstat \

unixODBC \

unixODBC-devel \

xorg-x11-xauth \

xorg-x11-utils

8. Edit the sysctl file. vim /etc/sysctl.conf

fs.file-max = 6815744

kernel.sem = 250 32000 100 128

kernel.shmmni = 4096

kernel.shmall = 1073741824

kernel.shmmax = 4398046511104

net.core.rmem_default = 262144

net.core.rmem_max = 4194304

net.core.wmem_default = 262144

net.core.wmem_max = 1048576

fs.aio-max-nr = 1048576

net.ipv4.ip_local_port_range = 9000 65500

vm.nr_hugepages = 52248

vm.hugetlb_shm_group = 54321

9. Apply the changes with the following command: sysctl -p

10. Edit the security limits configuration. vim /etc/security/limits.conf

oracle soft nofile 1024

oracle hard nofile 65536

oracle soft nproc 2047

oracle hard nproc 16384

oracle soft stack 10240

oracle hard stack 32768

grid - nofile 65536

grid - nproc 16384

grid - stack 32768

oracle soft memlock 536870912

oracle hard memlock 536870912

11. Add the necessary groups and users. groupadd -g 1001 oinstall

groupadd -g 1002 dba

groupadd -g 1003 asmadmin

groupadd -g 1004 asmdba

useradd -m -u 1002 -g oinstall -G dba,asmadmin,asmdba oracle

useradd -m -u 1003 -g oinstall -G dba,asmadmin,asmdba grid

12. Add the following lines to the .bash_profile file for the oracle user: export TMP=/tmp

export TMPDIR=$TMP

export ORACLE_HOSTNAME=hostname

export ORACLE_UNQNAME=ORCL



export ORACLE_BASE=/u01/app/oracle

export GRID_HOME=/u01/app/12.1.0/grid

export DB_HOME=$ORACLE_BASE/product/12.1.0/dbhome_1

export ORACLE_HOME=$DB_HOME

export ORACLE_SID=orcl

export ORACLE_TERM=xterm

export BASE_PATH=/usr/sbin:$PATH

export PATH=$ORACLE_HOME/bin:$BASE_PATH

export LD_LIBRARY_PATH=$ORACLE_HOME/lib:/lib:/usr/lib

export CLASSPATH=$ORACLE_HOME/JRE:$ORACLE_HOME/jlib:$ORACLE_HOME/rdbms/jlib

13. Add the following lines to the .bash_profile file for the grid user: export TMP=/tmp

export TMPDIR=$TMP

export ORACLE_HOSTNAME=hostname

export ORACLE_BASE=/u01/app/grid

export GRID_HOME=/u01/app/12.1.0/grid

export ORACLE_HOME=$GRID_HOME

export ORACLE_SID=+ASM1

export ORACLE_TERM=xterm

export BASE_PATH=/usr/sbin:$PATH

export PATH=$ORACLE_HOME/bin:$BASE_PATH

export LD_LIBRARY_PATH=$ORACLE_HOME/lib:/lib:/usr/lib

export CLASSPATH=$ORACLE_HOME/JRE:$ORACLE_HOME/jlib:$ORACLE_HOME/rdbms/jlib

14. Create the following directories, and assign the following permissions. mkdir -p /u01/app/oracle

mkdir /u01/app/grid

chown -R oracle:oinstall /u01/app

chmod -R g+w /u01/app

15. Create passwords for the oracle and grid accounts with passwd.

16. Edit the hosts file. vim /etc/hosts

127.0.0.1 R720 R720.localhost.localdomain localhost

localhost.localdomain localhost4 localhost4.localdomain4

::1 R720 R720.localhost.localdomain localhost

localhost.localdomain localhost6 localhost6.localdomain6

17. Edit the /etc/fstab file. vim /etc/fstab

Adding this line:

tmpf /dev/shm tmpfs defaults,size=89G 0 0

18. Edit the 90-nproc.conf file. vim /etc/security/limits.d/90-nproc.conf

Modifying this line:

* soft nproc 1024

To reflect this change:



* - nproc 16384

19. Download IONTuner from Fusion-IO and install with rpm -i iontuner_1.1.0_1.el6.noarch.rpm.

This is available to download from www.fusionio.com/files/ion-optimization-scripts.

20. Edit the /boot/grub/menu.lst file:

vim /boot/grub/menu.lst

Append this to the kernel line: intel_idle.max_cstate=0 processor.max_cstate=0

21. Enable the multipath configuration file and start the multipath daemon:

mpathconf --enable --with_multipathd y.

22. For the DAAD configurations, edit /etc/multipath.conf to add the following:

vim /etc/multipath.conf

defaults {

user_friendly_names yes

queue_without_daemon no

}

devices {

device {

vendor "FUSIONIO"

features " 3 queue_if_no_path pg_init_retries 50" hardware_handler "1 alua"

path_grouping_policy group_by_prio

path_selector "queue-length 0"

failback immediate

path_checker tur

prio alua

fast_io_fail_tmo 15

dev_loss_tmo 60

}

device {

vendor "IBM"

product "FlashSystem-9840"

path_selector "round-robin 0"

path_grouping_policy multibus

path_checker tur

rr_min_io_rq 4

rr_weight uniform

no_path_retry fail

dev_loss_tmo 300

fast_io_fail_tmo 25

}

}

23. Perform a Loop Initialization Protocol and rescan the Fibre Channel interconnects with echo "1" >

/sys/class/fc_host/hostX/issue_lip, replacing X with each of the hosts being used on the system.

24. Restart the multipath service with service multipathd restart.

25. Verify that multipath is working with multipath -ll.

http://www.fusionio.com/files/ion-optimization-scripts



26. Create a primary partition with the fdisk utility on each multipath disk with the following:

a. fdisk /dev/mapper/mpathX, where X is the identifier of the disk in question.

b. n

c. p

d. 1

e. 1

f. Press Enter.

g. w

27. Reboot the server to make the kernel use the new table.

28. Edit the scsi_id file. echo "options=-g" > /etc/scsi_id.config

Presenting DAAD and IBM LUNs to Oracle ASM We used the steps below to configure the multipathed storage LUNs prior to setting up ASM.

1. Edit the 99-oracle-asmdevices rules file. vim /etc/udev/rules.d/99-oracle-asmdevices.rules

KERNEL=="dm-*", ENV{DM_UUID}=="mpath-26230373534643030",

SYMLINK+="oracleasm/DAADdata1", OWNER="oracle", GROUP="dba", MODE="0660"












SYMLINK+="oracleasm/DAADlogs1", OWNER="oracle", GROUP="dba", MODE="0660"


SYMLINK+="oracleasm/DAADlogs2", OWNER="oracle", GROUP="dba", MODE="0660"

KERNEL=="dm-*", ENV{DM_UUID}=="mpath-360a980004431455a345d4733696e6937",

SYMLINK+="oracleasm/IBMdata1", OWNER="oracle", GROUP="dba", MODE="0660"








SYMLINK+="oracleasm/IBMlogs1", OWNER="oracle", GROUP="dba", MODE="0660"





KERNEL=="dm-*", ENV{DM_UUID}=="mpath-360a980004431455a345d4733696e694b",




KERNEL=="dm-*", ENV{DM_UUID}=="mpath-360a980004431455a345d4733696e694d",

SYMLINK+="oracleasm/IBMbackup1", OWNER="oracle", GROUP="dba", MODE="0660"

KERNEL=="dm-*", ENV{DM_UUID}=="mpath-360a980004431455a345d4733696e694f",

SYMLINK+="oracleasm/IBMbackup2", OWNER="oracle", GROUP="dba", MODE="0660"

2. Execute udevadm and start udev. udevadm control --reload-rules

start_udev

3. List the ASM devices. ls -l /dev/oracleasm/

lrwxrwxrwx 1 root root 8 Mar 20 11:29 DAADdata1 -> ../dm-13






lrwxrwxrwx 1 root root 8 Mar 20 11:29 DAADlogs1 -> ../dm-16

lrwxrwxrwx 1 root root 8 Mar 20 11:29 DAADlogs2 -> ../dm-22

lrwxrwxrwx 1 root root 8 Mar 20 11:29 IBMbackup1 -> ../dm-10

lrwxrwxrwx 1 root root 8 Mar 20 11:29 IBMbackup2 -> ../dm-11

lrwxrwxrwx 1 root root 7 Mar 20 11:29 IBMdata1 -> ../dm-3




lrwxrwxrwx 1 root root 7 Mar 20 11:29 IBMlogs1 -> ../dm-7




Installing Oracle Grid Infrastructure for Standalone Server 12c Prior to starting the steps below, we downloaded the Oracle 12c Grid installation and extracted it to the /grid

directory.

1. Run the GUI installer for Oracle Database using the following commands:

ssh –Y grid@R720_IP_address

cd /grid

./runInstaller

2. Launch the Oracle Grid Infrastructure installation wizard.

3. In Software Updates, select Skip software updates, and click Next.

4. In Installation Options, select Install and Configure Oracle Grid Infrastructure for a Standalone Server, and click Next.

5. In Product Languages, select English and click the right-pointing arrow between the two selection panels to add

English to the Selected languages panel. Click Next.

6. In Create ASM Disk Group, set the Disk group name to DATA.

7. Click Change Discovery Path.

8. Enter /dev/oracleasm for the Disk Discovery Path, and click OK.

9. Check the boxes for the appropriate data disks, and Click Next.

a. For the DAAD-only configuration, select only the DAADs’ data disks.



b. For the IBM-only configuration, select only the IBM’s data disks.

10. In ASM Password, select Use same passwords for these accounts. Enter and confirm the password, and click Next.

11. In Operating System Groups, set Oracle ASM Administrator Group to asmadmin, and Oracle ASM DBA Group to

asmdba, and click Next.

12. In Installation Location, accept the default locations provided, and click Next.

13. In Create Inventory, accept the defaults, and click Next.

14. In Root Script Execution, check the box for Automatically run configuration scripts.

15. Select Use “root” user credential, and provide the root password. Click Next.

16. In Summary, review the information, and click Install to begin installation.

17. Click Yes to confirm using the privileged user for the installer.

18. In Finish, click Close to exit the installer.

Installing Oracle Database 12c Prior to starting the steps below, we downloaded the Oracle Database 12c installation and extracted it to the

/database directory.

1. Run the GUI installer for Oracle Database using the following commands:

ssh –Y oracle@R720_IP_address

cd /database

./runInstaller

2. Launch the Oracle Database 12c Release 1 Installer.

3. In Configure Security Updates, uncheck the I wish to receive security updates via My Oracle Support checkbox. Click

Next.

4. Click Yes to confirm no email provided, and continue.


6. In Installation Options, select Install database software only, and click Next.

7. In Grid Installation Options, select Single instance database installation, and click Next.

8. In Product Languages, select English and click the right-pointing arrow located between the two selection panels to

add English to the Selected languages panel. Click Next.

9. In Database Edition, select Enterprise Edition, and click Next.

10. In Installation Location, accept the default locations provided, and click Next.

11. In Operating System Groups, accept the defaults, and click Next.


13. When prompted, follow the instructions to execute the scripts. Click OK when the scripts have completed.


Creating Oracle ASM disk groups for the database 1. Log into the system as the grid user.

2. Start the ASM configuration assistant, asmca.

3. On the Disk Groups tab, click Create.

4. On the Create Disk Group pop-up screen, enter LOGS for the Disk Group Name.



5. Select External (None) for Redundancy.

6. Select /dev/oracleasm/ for the Disk Path.

7. Select the two logs disks.

a. To start with the DAAD-only configuration, select only the DAADs’ logs disks.

b. When switching to another configuration later, launch asmca again, and perform the Drop Disks and Add

Disks operations as necessary to achieve the correct arrangements,

i. For the IBM configuration, the DATA disk group should contain the IBM data disks, the LOGS disk

group should contain the IBM logs disks, and the BACKUP disk group should contain the IBM backup

disks.

ii. For the DAAD configuration, the DATA disk group should contain six of the DAAD disks, the LOGS

disk group should contain other two DAAD disks, and the BACKUP disk group should still contain the

IBM backup disks.

8. Click OK to create the LOGS disk group, and click OK on the completed-task pop-up screen

9. Repeat steps 4 through 9 to create a BACKUP disk group with the IBM backup disks.

10. Click Exit to close the ASM configuration assistant.

Creating the Oracle Database (using DBCA) 1. Log into the system as the Oracle user.

2. Launch the Database Configuration Assistant (DBCA).

3. In Database Operations, select Create Database, and click Next.

4. In Creation Mode, select Advanced Mode, and click Next.

5. In Database Template, select the Template for General Purpose or Transaction Processing, and click Next.

6. If prompted about continuing without upgrade, click Yes.

7. In Database Identification, type orcl for the Global Database Name.

8. Type orcl for the SID. Click Next.

9. In Management Options, select Configure Enterprise Manager (EM) Database Express. Click Next.

10. In Database Credentials, select Use the Same Administrative Password for All Accounts.

11. Enter and confirm the administrative password, and click Next.

12. In Network Configuration, check the boxes for all listeners, and click Next.

13. In Storage Locations, select ASM for Database Storage Type.

14. Select User Common Location for All Database Files, and type +DATA into the Database Files Location field.

15. Select ASM for Recovery files Storage Type.

16. Specify Fast Recovery Area. Type +BACKUP in the Fast Recovery Area field.

17. Set the Fast Recovery Area size to 1850 GB, and click Next.

18. In Database Options, accept the defaults, and click Next.

19. In Initialization Parameters and under typical settings, set the Memory Size to 40%, and click next.

20. In Creation Options, select Create Database.

21. Click Next.

22. Review the Summary. To complete the database creation, click Finish.



23. Review the information on the screen, and click Exit.

24. To exit the DBCA, click Close.

Configuring the Oracle Tablespace and redo logs Alter the tablespace and redo logs as follows:

ALTER DATABASE ADD LOGFILE GROUP 4 ( '/tmp/temp1.log' ) SIZE 50M;

ALTER DATABASE ADD LOGFILE GROUP 5 ( '/tmp/temp2.log' ) SIZE 50M;

ALTER SYSTEM SWITCH LOGFILE;


ALTER SYSTEM CHECKPOINT;

ALTER DATABASE DROP LOGFILE GROUP 1;









-- DELETE LOGS IN ASM HERE --

ALTER DATABASE ADD LOGFILE GROUP 1 ( '+LOGS/ORCL/ONLINELOG/redo01.log' ) SIZE

20G;


20G;


20G;











HOST rm -f /tmp/temp*.log

CREATE BIGFILE TABLESPACE "TPCC"



DATAFILE '+DATA/orcl/tpcc.dbf' SIZE 400G AUTOEXTEND ON NEXT 1G

BLOCKSIZE 8K

EXTENT MANAGEMENT LOCAL AUTOALLOCATE

SEGMENT SPACE MANAGEMENT AUTO;

CREATE BIGFILE TABLESPACE "TPCC_OL"

DATAFILE '+DATA/orcl/tpcc_ol.dbf' SIZE 150G AUTOEXTEND ON NEXT 1G

BLOCKSIZE 16K

EXTENT MANAGEMENT LOCAL AUTOALLOCATE

SEGMENT SPACE MANAGEMENT AUTO;

ALTER DATABASE DATAFILE '+DATA/ORCL/DATAFILE/UNDOTBS1.260.873465895' RESIZE

32760M;

Configuring the Oracle pfile Alter the Oracle pfile as follows:

orcl.__oracle_base='/u01/app/oracle'#ORACLE_BASE set from environment

_enable_NUMA_support=TRUE

_kgl_hot_object_copies=2

_shared_io_pool_size=512m

audit_file_dest='/u01/app/oracle/admin/orcl/adump'

audit_trail='NONE'

compatible='12.1.0.2.0'

control_files='+DATA/ORCL/CONTROLFILE/current.261.873465981','+BACKUP/ORCL/CONTRO

LFILE/current.256.873465981'#Restore Controlfile

db_16k_cache_size=16g

db_block_size=8192

db_cache_size=64g

db_create_file_dest='+DATA'

db_domain=''

db_name='orcl'

db_recovery_file_dest_size=1850g

db_recovery_file_dest='+BACKUP'

diagnostic_dest='/u01/app/oracle'

disk_asynch_io=TRUE

dispatchers='(PROTOCOL=TCP) (SERVICE=orclXDB)'

dml_locks=500

fast_start_mttr_target=300

java_pool_size=4g

job_queue_processes=0

large_pool_size=1g

local_listener='LISTENER_ORCL'

lock_sga=TRUE

log_buffer=536870912#log buffer update

log_checkpoint_interval=0

log_checkpoint_timeout=0

log_checkpoints_to_alert=TRUE

open_cursors=3000

parallel_max_servers=0

pga_aggregate_target=6g

recovery_parallelism=4



remote_login_passwordfile='EXCLUSIVE'

resource_manager_plan='FORCE'

shared_pool_size=3g

undo_management='AUTO'

undo_retention=1

undo_tablespace='UNDOTBS1'

use_large_pages='ONLY'

Setting up the HammerDB client We used a dual-socket server running Red Hat Enterprise Linux 6.6 for the HammerDB client. We followed the

installation steps at the beginning of this appendix to install Red Hat Enterprise Linux, and also installed the GUI. We

then installed the HammerDB client software.

Installing HammerDB Download and install version 2.16 on the Red Hat client. We downloaded HammerDB from the following

location: hammerora.sourceforge.net/download.html. We installed HammerDB according to the installation guide

(hammerora.sourceforge.net/hammerdb_install_guide.pdf).

Installing HammerDB Oracle libraries Complete the following steps on the application server.

1. Launch the Oracle Client Installer.

2. In Select Installation Type, select Administrator (1.8 GB) as the installation type, and click Next.


4. In Select Product Languages, select English and click the right-pointing arrow located between the two selection

panels to add English to the Selected languages panel. Click Next.

5. In Specify Installation Location, accept the default locations provided, and click Next.

6. In Create Inventory, accept the defaults, and click Next.


8. In Install Product, follow the instructions to execute the scripts. Click OK when the scripts have completed.


Configuring the database We used the TPC-C build schema build options for Oracle inside HammerDB to build the database. We set the

following options in the build schema.

Oracle Service Name = R720_IP_address/orcl

System User = SYSTEM

System User Password = Password1

TPC-C User = tpcc

TPC-C User Password = tpcc

TPC-C Default Tablespace = tpcc

Order Line Tablespace = tpcc_ol

TPC-C Temporary Tablespace = temp

TimesTen Database Compatible = unchecked

Partition Order Line Table = checked

Number of Warehouses = 5000

Virtual Users to Build Schema = 20

http://hammerora.sourceforge.net/download.htmlhttp://hammerora.sourceforge.net/hammerdb_install_guide.pdf



Use PL/SQL Server Side Load = unchecked

Server Side Log Directory = /tmp

Running HammerDB We ran HammerDB by filling in the appropriate information for the driver options. We tested with 10,000,000

transactions per user, a 20-minute ramp up time and 20-minute test duration. We used 101 virtual users with 0-ms user

delay and repeat delay. We used rman to back up the database before testing, and we ran restores between runs.



ABOUT PRINCIPLED TECHNOLOGIES

Principled Technologies, Inc. 1007 Slater Road, Suite 300 Durham, NC, 27703 www.principledtechnologies.com

We provide industry-leading technology assessment and fact-based marketing services. We bring to every assignment extensive experience with and expertise in all aspects of technology testing and analysis, from researching new technologies, to developing new methodologies, to testing with existing and new tools. When the assessment is complete, we know how to present the results to a broad range of target audiences. We provide our clients with the materials they need, from market-focused data to use in their own collateral to custom sales aids, such as test reports, performance assessments, and white papers. Every document reflects the results of our trusted independent analysis. We provide customized services that focus on our clients’ individual requirements. Whether the technology involves hardware, software, Web sites, or services, we offer the experience, expertise, and tools to help our clients assess how it will fare against its competition, its performance, its market readiness, and its quality and reliability. Our founders, Mark L. Van Name and Bill Catchings, have worked together in technology assessment for over 20 years. As journalists, they published over a thousand articles on a wide array of technology subjects. They created and led the Ziff-Davis Benchmark Operation, which developed such industry-standard benchmarks as Ziff Davis Media’s Winstone and WebBench. They founded and led eTesting Labs, and after the acquisition of that company by Lionbridge Technologies were the head and CTO of VeriTest.

Principled Technologies is a registered trademark of Principled Technologies, Inc. All other product names are the trademarks of their respective owners.

Disclaimer of Warranties; Limitation of Liability: PRINCIPLED TECHNOLOGIES, INC. HAS MADE REASONABLE EFFORTS TO ENSURE THE ACCURACY AND VALIDITY OF ITS TESTING, HOWEVER, PRINCIPLED TECHNOLOGIES, INC. SPECIFICALLY DISCLAIMS ANY WARRANTY, EXPRESSED OR IMPLIED, RELATING TO THE TEST RESULTS AND ANALYSIS, THEIR ACCURACY, COMPLETENESS OR QUALITY, INCLUDING ANY IMPLIED WARRANTY OF FITNESS FOR ANY PARTICULAR PURPOSE. ALL PERSONS OR ENTITIES RELYING ON THE RESULTS OF ANY TESTING DO SO AT THEIR OWN RISK, AND AGREE THAT PRINCIPLED TECHNOLOGIES, INC., ITS EMPLOYEES AND ITS SUBCONTRACTORS SHALL HAVE NO LIABILITY WHATSOEVER FROM ANY CLAIM OF LOSS OR DAMAGE ON ACCOUNT OF ANY ALLEGED ERROR OR DEFECT IN ANY TESTING PROCEDURE OR RESULT. IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC. BE LIABLE FOR INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES IN CONNECTION WITH ITS TESTING, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. IN NO EVENT SHALL PRINCIPLED TECHNOLOGIES, INC.’S LIABILITY, INCLUDING FOR DIRECT DAMAGES, EXCEED THE AMOUNTS PAID IN CONNECTION WITH PRINCIPLED TECHNOLOGIES, INC.’S TESTING. CUSTOMER’S SOLE AND EXCLUSIVE REMEDIES ARE AS SET FORTH HEREIN.
http://www.principledtechnologies.com

Documents

Dell Acceleration Appliance for Databases vs. IBM ... › Dell › DAAD_vs_IBM_0815.pdfDell Acceleration Appliance for Databases vs. IBM FlashSystem 900: Capacity, cost, and performance